Notes 16a

Transcription

Notes 16a
Week 11: Chap. 16a Pulse Processing
Fast Neutron Detection
Pulse Processing (passive)
-- Signal shape
Big RIPS Electronics
-- Cable properties RIKEN, October, 2o11
--- connecters
--- impedance
-- CR, RC filters
Pulse Processing
(active)
© DJMorrissey, 2o15
Pulse Processing: overview
Fig. 16.1 Knoll, 4th Ed.
© DJMorrissey, 2o15
Fig. 16.2 Knoll, 4th Ed.
Chap. 16a – Pulse Processing
Fig. 4.1 Knoll, 3rd , 4th Eds.
or “DROOP”
Leading or Falling edge
Rise or Fall time – 10:90%
(Uni or Bipolar // Analog or Digital)
© DJMorrissey, 2o15
Fig. 11.1 Leo, 2nd Ed.
Pulse Processing: cables
Twisted pairs – “differential” signals, logic or analogue
Coaxial conductor/shield – signal on the “center”
Each configuration has a RG/U name with a
characteristic capacitance and inductance per
unit length, and an impedance (with a
negligible resistance).
d 2V
= −ω 2 L C V
0
0
dx 2
V = Ae jkx + Be − jkx
v=
ω
k
Z0 =
© DJMorrissey, 2o15
=
L0
C0
1
L0C0
k 2 = ω 2 L0C0
for coax
C0 = 2πε /ln(r2 /r1 )
L0 = (µ /2π ) ln(r2 /r1 )
Pulse Processing: some connectors
CD International Technology
http://www.cdint.com/catalog/model/CC-B
N
“BNC is an acronym for Bayonet Neill-Concelman, after Paul Neill of Bell Labs (inventor
of the N connector) and Amphenol engineer Carl Concelman (inventor of the C connector).
BNC is often erroneously expanded to "Baby Neill-Concelman", "Baby N connector",
"British Naval Connector", and "Bayonet Nut Connector".”
“This connector has a characteristic impedance of 50 ohms, and needs to be mated with 50
ohm coaxial cable in order to prevent signal loss, noise, and/or transmitter damage due to
signal reflections at the point of mismatch. 50 ohm coax cable, connectors, and adapters are
commonly used in wifi (802.11 wireless LAN) antennas, ham tranceivers, and other radio
frequency (RF) analog and digital signaling, microwave, radar, hi-fidelity professional
audio, non-destructive testing (NDT), oil and petroleum production, ultrasonic transducers,
accelerometers, strain gauges, and some professional video applications.”
BNC
C
LEMO documentation
http://www.lemo.com/en/documentation
(Léon Mouttet)
LEMO to
© DJMorrissey, 2o15
BNC
Pulse Processing: impedance matching
Fig. 16.4 Knoll, 3rd Ed.
16.7, 4th Ed.
I Inc + I refl = I L
Z0
VInc +Vrefl = VL
Vrefl = −I refl Z 0
ZL
→
What happens if: ?
Vrefl
• Open circuit .. ZL ~ ∞
• Short circuit .. ZL = 0
• Match circuit .. ZL = Z0
VInc
Vrefl Z L − Z 0
=
VInc Z L + Z 0
VL
+1
2
−1
0
0
1
VL
2Z L
=
VInc Z L + Z 0
VInc
Match to get maximum transmission to load (S/N), minimize
reflections (ringing), maintain signal shape.
Don’t Match to minimize transmission .. Weak signals into high
impedance loads or low power sources (but must use short cables).
© DJMorrissey, 2o15
Pulse Processing: simplest manipulations
Voltage divider (recall PMT base)
Attenuation α = Vout/Vin
R1=R0 α-1 / α +1
R2=R0 2 α / α2-1
Fig. 16.5 Knoll, 3rd Ed.
16.8, 4th Ed.
Signal Inverter
Signal splitter: R=R0/3
Fig. 16.7 Knoll, 3rd Ed.
16.10 4th Ed.
Fig. 16.6 Knoll, 3rd Ed
16.9, 4th Ed..
© DJMorrissey, 2o15
All loads must be present, or else see
previous discussion about reflections!
Pulse Processing: CR circuit
Fig. 16.9 Knoll, 3rd Ed.
17.2
4th
Ed.
Vout ≈ τ
€
Q
+Vout
(C is constant)
C
dVin 1 dQ dVout
dQ
=
+
=i =V / R
dt
C dt
dt
dt
dVin Vout dVout
=
+
τ = RC
dt
RC
dt
dV
dV
τ in = Vout + τ out
dt
dt
Vin =
dVin
dt
dVout
dV
≈ τ in
dt
dt
Vout ≈ Vin
τ
for small τ
for large τ
The differentiator (when τ is small, “fast” electronics)
.. Should remove low frequency components and is called a
“high-pass” filter.
© DJMorrissey, 2o15
Pulse Processing: RC circuit
Fig. 16.9 Knoll, 3rd Ed.
17.3, 4th Ed.
Vin = iR +Vout
dQ
dV
= C out (on capacitor)
dt
dt
dV
Vin = RC out +Vout
dt
dV
Vin = τ out +Vout
dt
i=
Vin
dVout
≈
τ
dt
→
1
V
∫
τ
in
dt ≈ Vout
Vout ≈ Vin
for small RC
€
The integrator (when τ is large, “slow”
electronics)
.. Should remove high frequency components and is called a
“low-pass” filter.
© DJMorrissey, 2o15
Pulse Processing: Cable Properties
v=
ω
1
=
k
L0C0
Z0 =
L0
C0
so that
© DJMorrissey, 2o15
Z 0C0 =
1
= L0C0 = "τ "
v
However, this is an unusual time constant –
note that it has dimensions of (time / length )
The τ depends on the length of the cable.
Some example properties from Belden Cables:
58 /U
59 /U
213/U
316 /U
Zo
50 Ω
75
50
50
Co
24.3 pF/ft
16.3
30.8
29
Lo
0.064 µH/ft
0.107
0.077
0.067
v/c
0.77
0.83
0.66
0.695
τ
1.2 ns/ft
(4 ns/m)
1.3 ns/ft 1.5 ns/ft 1.4 ns/ft
Pulse Processing: Cable Consequences
v=
ω
1
=
k
L0C0
L0
C0
so that
Z 0C0 =
1
= L0C0 = "τ "
v
However, this is an unusual time constant –
note that it has dimensions of (time / length )
The τ depends on the length of the cable.
Some example properties from Belden Cables:
Attenuation (dB/100m)
Z0 =
Fig. 13.10 Leo, 2nd Ed.
© DJMorrissey, 2o15
1000$
58/U$
59/U$
213/U$
100$
316/U$
10$
1$
1$
0.1$
10$
100$
1000$
Frequency (MHz)
10000$
Pulse Processing: Question
Electronics in the vault or not?
Compare the output of a preamp step-function pulse that passes through a 1m
Beldin RG-58/U cable to that that passing through 50 m of the same cable. Use
the Fermi function with a=1 , t0=10, τ = 4 ns/m, and t in ns.
f (t ) = 1 /(1 + e −(t −to) / a )
© DJMorrissey, 2o15
Pulse Processing: Answer
Electronics in the vault or not?
Compare the output of a preamp step-function pulse that passes through a 1m
Beldin RG-58/U cable to that that passing through 50 m of the same cable. Use
the Fermi function with a=1 , t0=10, τ = 4 ns/m, and t in ns.
f (t ) = 1 /(1 + e −(t −to) / a )
1.0
1.0
RC : 1 / (1+ e−(t−to)/a ) [1− e−(t−to)/τ ]
0.8
0.8
0.6
0.6
CR : (1 / (1+ e−(t−to)/a ))2 e−(t−to)/τ
0.4
0.4
0.2
0.2
CR & RC : (1 / (1+ e−(t−to)/a ))2 e−(t−to)/τ [1− e−(t−to)/τ ]
00
100
100
200
200
© DJMorrissey, 2o15
300
300
400
400
500
500
600
600
700
700